set plugin to use local api

[clude]

do you have a plan to make the vs/jetbrains plugin supporting local deployed CodeGeeX Service

[ImJoyed]

The latest plugins are now supported.

[clude]

cooooooool!!!!
found the latest VS plugin working on local ai model now! will the next released jetbrains plugin have the same feature ?

[ShaoboZhang]

cooooooool!!!! found the latest VS plugin working on local ai model now! will the next released jetbrains plugin have the same feature ?

Yes, but Jetbrains has a stricter review process, so it takes a bit longer than vs code.

[agdkgg]

ollama 本地 连接失败 403错误代码，网络 后缀名试了几个，是代码还没修改吗？

[limingchina]

I've seen those settings in vscode

 "Codegeex.Local": {

        "apiURL": "",
        "useChatGLM": true,
        "chatGLM": {
            "apiKey": "",
            "model": ""
        },

What should be set for them? @clude , have you tried it successfully?

[clude]

you just need to config the openai api address and model name。 by the way , if u use ollama as a service , don't forget to set the env value: OLLAMA_ORIGINS="*"

[Choumingzhao]

Got same error with codegeex vscode plugin.
Using Ollama api: http://localhost:11434/api/generate. Ollama 0.2.1 API test passed with curl both on local and remote. I also tried other ollama code completion plugins in vscode and they worked.

The screenshot of local api usage by @Stanislas0 in local mode tutorials only show the openai style api . So ollama 0.2.1 api (api/generate, api/chat) may not work in current plugin. Edit: tested openai style api and still get same network error.

Output from CodeGeeX Local in VSCode:

• Ollama style api:

  
  [info] [Model Config] try to connect: http://localhost:11434/api/generate

data: { "messages": [ { "role": "user", "content": "Hi" } ], "model": "codegeex4", "stream": false, "temperature": 0.9, "top_p": 0.9, "stop": [] } headers: {} [error] [Model Config] error: { "message": "Network Error", "name": "AxiosError", "stack": "AxiosError: Network Error\n at u.onerror (https://vscode-remote+ssh-002dremote-002bjss.vscode-resource.vscode-cdn.net/home/user/.vscode-server/extensions/aminer.codegeex-2.12.2/webview-ui/dist/assets/index.js:55:1051909)\n at XMLHttpRequest.r (https://vscode-remote+ssh-002dremote-002bjss.vscode-resource.vscode-cdn.net/home/user/.vscode-server/extensions/aminer.codegeex-2.12.2/webview-ui/dist/assets/index.js:35:854177)", "config": { "transitional": { "silentJSONParsing": true, "forcedJSONParsing": true, "clarifyTimeoutError": false }, "adapter": ["xhr", "http"], "transformRequest": [null], "transformResponse": [null], "timeout": 0, "xsrfCookieName": "XSRF-TOKEN", "xsrfHeaderName": "X-XSRF-TOKEN", "maxContentLength": -1, "maxBodyLength": -1, "env": {}, "headers": { "Accept": "application/json, text/plain, /", "Content-Type": "application/json", "Exp-Id": "", "Variant-Id": "", "Is-Exp-Active": "false" }, "baseURL": "https://codegeex.cn/prod", "method": "post", "url": "http://localhost:11434/api/generate", "data": "{\"messages\":[{\"role\":\"user\",\"content\":\"Hi\"}],\"model\":\"codegeex4\",\"stream\":false,\"temperature\":0.9,\"top_p\":0.9,\"stop\":[]}" }, "code": "ERR_NETWORK", "status": null }

* OpenAI style compatible API also by ollama
```quote
2024-07-16 10:12:45.105 [info] [Model Config] try to connect: http://localhost:11434/v1/chat/completions 

  data: {"messages":[{"role":"user","content":"Hi"}],"model":"codegeex4","stream":false,"temperature":0.9,"top_p":0.9,"stop":[]}
  headers: {}

2024-07-16 10:12:45.144 [error] [Model Config] error: {"message":"Network Error","name":"AxiosError","stack":"AxiosError: Network Error\n    at u.onerror (https://vscode-remote+ssh-002dremote-002bjss.vscode-resource.vscode-cdn.net/home/user/.vscode-server/extensions/aminer.codegeex-2.12.2/webview-ui/dist/assets/index.js:55:1051909)\n    at XMLHttpRequest.r (https://vscode-remote+ssh-002dremote-002bjss.vscode-resource.vscode-cdn.net/home/user/.vscode-server/extensions/aminer.codegeex-2.12.2/webview-ui/dist/assets/index.js:35:854177)","config":{"transitional":{"silentJSONParsing":true,"forcedJSONParsing":true,"clarifyTimeoutError":false},"adapter":["xhr","http"],"transformRequest":[null],"transformResponse":[null],"timeout":0,"xsrfCookieName":"XSRF-TOKEN","xsrfHeaderName":"X-XSRF-TOKEN","maxContentLength":-1,"maxBodyLength":-1,"env":{},"headers":{"Accept":"application/json, text/plain, */*","Content-Type":"application/json"},"baseURL":"https://codegeex.cn/prod","method":"post","url":"http://localhost:11434/v1/chat/completions","data":"{\"messages\":[{\"role\":\"user\",\"content\":\"Hi\"}],\"model\":\"codegeex4\",\"stream\":false,\"temperature\":0.9,\"top_p\":0.9,\"stop\":[]}"},"code":"ERR_NETWORK","status":null}

For those who are eager to try CodeGeeX4 model in Ollama now, I recommand Continue. I managed to use the model with the plugin.

[clude]

For me , I tried CodeQwen1.5 on ollama 0.1.44 and CodeGeex4-9b on Ollama 0.2.5, both are working。

currently, if VsCode is your primary IDE , "Continue" may be the best choice, it is stable、customizable and has much more features than CodeGeex plugin in local mode. but "Continue"'s jetbrains plugin will disappoint you， it is not stable, always crash。

so, I still have high hopes on the CodeGeex plugin's local model, hope it has the full feature as the online mode version.